基于人工神经网络算法的发射药钝感剂浓度分布快速预测技术

提出一种适用于5/7改性单基发射药的快速预测钝感剂浓度分布情况的方法,基于人工神经网络算法构建了神经网络模型,然后使用已有的5/7改性单基发射药钝感剂浓度分布试验数据训练模型。结果发现,经过训练后模型输出的钝感剂浓度分布曲线和试验测定的钝感剂浓度分布曲线二者之间的复相关系数R高达0.93。这说明构建的模型可以较准确地快速预测出给定工艺参数条件下5/7改性单基发射药钝感剂浓度分布曲线,相较于传统的测试方法,具有省时、省力、方便快捷的优势,证明此方法具有很好的实用性。基于此模型还可以反向应用,根据需要的钝感剂浓度分布曲线来预测所需要的钝感发射药制备存贮工艺条件。

Prediction of Desensitizer Concentration Distribution in Gun Propellant Based on Artificial Neural Network Algorithm

A method for quickly predicting the concentration distribution of desensitizer in 5/7 modified single base gun propellant was proposed.A simple neural network model based on artificial neural network algorithm was established,and the experimental data of desensitizer concentration profile of 5/7 modified single-base propellant was used to train the model.The results show that after data training,the complex correlation coefficient R between the desensitizer concentration distribution curve output by the model and the experiment reaches to 0.93.This result shows that the constructed model can accurately and quickly predict the concentration profile curve of desensitizer in 5/7 modified single-base gun propellant under the given process parameters.Compared with the traditional testing method,it has the advantages of saving time,labor and convenience,which proves that the method has certain practicability.The model can also be used to predict the required manufacturing and storage process conditions of desensitizer propellant according to the desired concentration profile of desensitizer.